Survey Research Design: Delivering Actionable Intelligence

Transcription

1 YOU WILL LEARN Survey Research Design: Delivering Actionable Intelligence Learn the foundations of actionable business intelligence through survey research design. CONTRIBUTING EDITOR Jared Heyman, Chairman, Infosurv Six simple steps to effective survey research design Types of research design, including exploratory, descriptive, and causal How Jergens and Mentadent used survey research design to deliver terrific tangible results A cost versus benefit analysis of survey research design Executive Summary How are you perceived among your employees, your customers, or your business partners? Are those perceptions grounded in solid fact or your best guess? Leading organizations turn opinions into information through solid survey research design. This white paper will introduce you to the foundations of the kind of actionable business intelligence that your organization can thrive on.

2 Table of Contents Solid Survey Research Design Helps Organizations Thrive... 3 What is Survey Research Design And Why Does it Matter?... 3 Types of Survey Research Design... 4 Exploratory Research... 4 Conclusive Research... 5 Descriptive Research... 6 Cross-Sectional vs. Longitudinal... 6 Causal Research... 6 A Solid Process For Effective Survey Research Design... 6 Six Essential Steps to Successful Research Design... 6 Navigating Through Sampling Error... 7 Cost versus Benefit Analysis... 7 Costs Associated with not Conducting Research... 8 Costs Associated with Conducting Research... 8 Bibliography... 9 Table of Figures... 9 Additional Insights Survey Research Design in Action: Mentadent... 4 What about Online Focus Groups?... 5 Survey Research Design in Action: Jergens... 5

3 Solid Survey Research Design Helps Organizations Thrive If you have a desire to develop real-world business intelligence, gain visibility, drive sales, improve employee relations, and better understand your market, then you need a well-executed and consistent process for designing and conducting your survey research. Once executed, effective survey research design can validate or disprove your perceptions about the business world around you an instrumental step in making important decisions. Through careful survey research design you can eliminate common problems and maximize the effectiveness of your results. Perception vs. Reality1 A company s perception A company that has true data A company that has no data or false data A company that thinks they have true data The perfect position A very dangerous position A company that thinks they have no data or poor data An opportunity to validate assumptions A true starting point for improvement Figure 1 What is survey research design and why does it matter? Survey research design is the blueprint for executing a marketing research project. It is the knowledge selection of the best research methods, often in combination, needed to gain information that will resolve management decisions. Research design defines the marketing research issue builds a conceptual framework denotes questions that need to be answered poses current hypotheses

4 defines information that must be obtained By utilizing various types of basic research designs, you can facilitate these important objectives. Types of Survey Research Design Research design traditionally falls into two main categories: exploratory and conclusive 4. Exploratory research builds insight into the problem at hand and develops working theories as to the cause and potential solutions. Conclusive research validates working theories and determines the best course of action in a given situation. Exploratory Research Since exploratory research is very fluid and unstructured, it is best employed when only a rough picture of the informational gaps that need to be filled exists. Goals of exploratory research include: Survey Research Design in Action: Mentadent Through survey research involving longitudinal track studies2, Mentadent Toothpaste learned they had strong brand loyalty. With this information, they decided to launch a new product, Mentadent Mouthwash. The company developed a strong advertising campaign capitalizing on the Mentadent brand and customer loyalty, and they tracked purchases and repeat purchases of their new product through a sample of panel households. These households provided information on purchases of toothpaste, mouthwash and other oral hygiene products, every month for a period of nine months. The results? Mentadent learned that their new product achieved a 3.5% dollar share of the category and built consistent growth in repeat purchases hitting 39% by the end of its introductory year 3. Identifying a business problem or defining it more precisely Identifying potential courses of action Building working theories Isolating key variables and relationships for further examination (breaking the issue into components) Gaining insight as to how to solve a specific business issue Categorizing priorities for further study

5 Here s an example: Sales are down, but the reasons are elusive. An exploratory research process helps in analyzing inbound calls to the call center to identify a possible pattern of complaints. Next, by establishing a set of focus groups comprised of existing The evolution of exploratory research In recent years, online and mobile technology platforms have begun to transform the world of exploratory research. It is now possible to host remote, moderated focus groups via webcam and collect video, image, and text data from mobile devices. The savvy researcher is tapping into these to reinforce or techniques to make their research more rich and efficient. customers, a vehicle for common complaints is established. From this process, it is possible for the emergence of new insight into the company's image, such as a need for more competitive pricing, or catering more to the improvement of customer service. Exploratory research typically samples a small non representative population, and the data obtained should be considered tentative. Descriptive research, then, is essential further validate information and disprove assumptions. Conclusive Research Conclusive research empowers the decision maker to evaluate and choose the best course of action in a given situation. It can be used to clarify, validate, refine or even discount lessons and insights gleaned from initial exploratory research. Conclusive research begins with the assumption that the researcher has an accurate grasp of the problem at hand, and the information needed to address this problem is clearly understood. Then, this assumption is tested and the specific relationships contained within are examined. Conclusive research is substantially more structured than exploratory research, with the process outlined prior to project kickoff. Typically, conclusive research has a much larger sample population size, and statistical techniques are used to analyze the resulting data. Survey Research Design in Action: Jergens The Andrew Jergens Company conducted a crosssectional survey to measure consumer perceptions, attitudes and use of soap bars. They discovered that users of the traditional soap bar did not like having to pick up the mushy bar, or having the leftover soap film in their shower or bath. And they did not like the dry feeling that was left on their skin. This data led Jergens to develop and launch a new product, Jergens Body Shampoo. This product has been so successful that it has revolutionized its category, and has in fact motivated the competition to launch ten new products of their own. Jergens Body shampoo is still the number one seller among body shampoos in food stores, and it has helped this category to grow to over $500 million by 2004 this is almost ¼ of the entire soap industry! 11

6 This type of research can be further broken down into two sub categories: descriptive and causal. Descriptive Research As the name implies, descriptive research aims to describe a market characteristic or function. As one of the most common forms of research, it can describe trends, identify relationships, or help to make future forecasts. Descriptive research consists of both cross-sectional designs and longitudinal designs. A cross-sectional design is a snapshot of the marketplace at a specific point in time where a selected group of respondents is measured only once. With a longitudinal design, changes in a sample population are measured over time by obtaining one or more measurements on the same variables during different timeframes. Cross-Sectional vs. Longitudinal 6 Cross-Sectional Design Large, representative sample Limited response bias Cost effective Longitudinal Design Ability to detect change Large amount of data collection Very accurate Figure 2 Causal Research Whereas descriptive research demonstrates market trends, causal research explains why those specific trends occur. A causal survey research design examines the cause and effect variables behind a hypothesis. Since it is conducted in a controlled environment, it results in a more accurate measurement of the causal variable. Traditional causal research often involves a laboratory environment, to control outside variables. This can make research costly and somewhat more complicated. However, recent developments in technology have moved this form of research out of the laboratory and into an online environment. A Solid Process for Effective Survey Research Design Survey research design should always be handled by professionals since relationships between exploratory, descriptive, and causal research can be interdependent, and the right balance must be struck in order to gather the right information. Consider this if a company finds themselves losing

7 market share, all three forms of research can be used to answer different key questions 7 : Exploratory Why are we losing market share? Descriptive How do consumers evaluate our brand as opposed to the competition? Causal How much will our sales increase if we decrease our price by 5%, 10% or 15%? Six essential steps to successful research design There are six essential steps to any successful survey research design process. 1. Define the information that is needed. 2. Design what type or types of research will best accomplish those goals. 3. Specify the measurement and scaling procedures for each type of research. 4. Build a questionnaire. 5. Specify the sampling process and the sampling size. 6. Build a strategy for data analysis. Navigating Through Sampling Error No research can give you 100% certainty or absolutely prove an assumption. There will always be what are known as random sampling and non-sampling errors. Random sampling errors occur when the sample includes an imperfect representation of the population of interest as a whole. Non-sampling errors include errors in the process itself, including problem definition, approach, questionnaire design, survey methods, data preparation and analysis. Sampling and non-sampling errors can be greatly reduced by investing in the right level of research, and in investing in the right sample population. Marketing research firms such as Infosurv can help you determine the size sample you need to confidently answer your research questions.

8 Cost Versus Benefit Analysis Any form of research will incur a cost to execute. What some organizations fail to realize, however, is that NOT conducting research also has a cost associated with it. The right investment in your research project is far more valuable than the potential cost of wrong business decisions. The rule to follow is that as the cost of failure increases, so should the sophistication and structure of your research. Consider the following questions as you formulate the potential ROI of your research investment. Costs Associated With Not Conducting Research Worst-case, how much could it cost us if we make the wrong decision? What risks, financial and otherwise, flow from having incomplete or inaccurate information? Costs Associated With Conducting Research What are the benefits, financial and otherwise, of knowing the answers to these questions? Can we put a value on the confidence and clarity that would result from clear information? What do we stand to save, or what risks do we minimize, by knowing more information? As you can see, there is a cost associated with NOT conducting research as well as a cost associated with conducting research. Survey research design is a matter of formulating the problem and designating a targeted research methodology to attack that problem. Whether the target is exploratory research, or a descriptive form of research to validate your assumptions, it s critical to design your survey research process in such a way that your research doesn t just deliver data it empowers you with valid, accurate information that you can turn into actionable knowledge. A variety of firms, including Infosurv, offer support to researchers interested in answering their questions in the most cost effective and accurate way possible. Consider inviting an experienced Infosurv research consultant to discuss your project with you today.

9 Bibliography 1 Based on concept created by Great Brook Consulting, 2 Longitudinal track studies covered in detail in this paper, pg Based on Oral Care, Progressive Grocer, 79(8), (August 2000): 18; Christine Brittar, Chesebrough, Colgate Stir Up Brush War, Brandweek, 40(14), (April 5, 1999): 4; 1995 Edison Best New Products Awards Winners, Marketing News, 30 (10), (May 6, 1996): E4-E11. 4 Concepts based on work created by Dr. Naresh K. Malhotra, Dr. Mark Peterson, Basic Marketing Research: A Decision- Making Approach: This information is based on direct observations made by Jared Heyman, President of Infosurv. 6 Based on work created by Dr. Naresh K. Malhotra, Dr. Mark Peterson, Basic Marketing Research: A Decision- Making Approach: Based on work created by Dr. Naresh K. Malhotra, Dr. Mark Peterson, Basic Marketing Research: A Decision- Making Approach: Based on work created by Dr. Naresh K. Malhotra, Dr. Mark Peterson, Basic Marketing Research: A Decision- Making Approach: For more information on this topic, see the Infosurv paper titled Questionnaire Design: An Expert Approach 10 For more information on this topic, see the Infosurv paper titled Survey Analysis Design: Keys to Making Sound Decisions from Survey Data 11 Based on Don Mills, Jergens Naturally Smooth Shave Minimizing Lotion Cosmetics, 30(5), (September 2002); Bath Boom Rah!, Discount Merchandiser, 39(6), (June 1999): 52; Rick Klein, 1995 Edison Best New Products Awards Winners, Marketing News, 30 (10), (May 6, 1996): E4-E11. Table of Figures Figure Figure